An inception‐ResNet deep learning approach to classify tumours in the ovary as benign and malignant

Abstract

The classification of tumours into benign and malignant continues to date to be a very relevant and significant research topic in the cancer research domain. With the advent of Computer Vision and rapid developments in the fields of deep learning, as well as medical devices and instruments, researchers are therefore utilizing the state‐of‐the‐art deep learning architectures to discover patterns in the medical image data and thereby use this information to detect tumours and classify them as benign or malignant. In this paper, we propose a custom state‐of‐the‐art deep learning architecture, the Inception‐ResNet v2 for the classification of ovarian tumours into the two categories of benign and malignant based on a custom dataset with a validation accuracy of 97.5% and a test accuracy of 67%. Furthermore, a quantum convolutional neural network (QCNN) was also implemented with an accuracy of 92% on the validation dataset.

Performance Analysis of Segmentation and Classification of CT-Scanned Ovarian Tumours Using U-Net and Deep Convolutional Neural Networks

“…After the region-based training in FaRe-ConvNN, a combination of SVC and Gaussian NB classifiers was used to classify the images, which resulted in impressive precision and recall values [ 17 ]. In the works carried out by Ashwini et al [ 18 , 19 , 20 ], various Deep Learning models were used to segment the CT scanned images and classify them using variants of CNN. In the work [ 18 , 19 ], Otsu’s method was used to segment the tumour and a dice score of 0.82 and Jaccard score of 0.8356 were obtained.…”

Performance Analysis of Segmentation and Classification of CT scanned Ovarian Tumours using UNet and Deep Convolutional Neural Networks

“…In the works carried out by Ashwini et al [18][19][20], the various deep learning models were used to segment the CT scanned images and classified using variants of CNN. In the work [18][19], the Otsu's method was used to segment the tumor and obtained the dice score of 0.82 and Jaccard score of 0.8356.…”

“…In the works carried out by Ashwini et al [18][19][20], the various deep learning models were used to segment the CT scanned images and classified using variants of CNN. In the work [18][19], the Otsu's method was used to segment the tumor and obtained the dice score of 0.82 and Jaccard score of 0.8356. Further the performance of the segmentation was used cGAN [20], in this study the segmentation and classification of tumors were carried out in the single pipeline and obtained the dice score of 0.91 and the Jaccard score of 0.89.…”

Semantic segmentation and classification of polycystic ovarian disease using attention UNet, Pyspark, and ensemble learning model

“…Using DL algorithms like the long short‐term memory model and CNN, they propose a model which will give ***more precise treatment for Ovarian tumours. Ashwini et al (Hegde & Kodipalli, 2022; Kodipalli & Devi, 2021; Kodipalli & Devi, 2023; Kodipalli, Devi, et al, 2022; Kodipalli, Fernandes, et al, 2023; Kodipalli, Guha, et al, 2022; Kodipalli, Gururaj, et al, 2023; Ruchitha et al, 2022), contributed extensively to the detection of PCOS using a questionnaire and found that Fuzzy TOPSIS outperformed the SVM algorithm (Kodipalli & Devi, 2021). Watershed and active contour random walker were used in (Ruchitha et al, 2022) for segmenting the ovarian tumour and it was found that the watershed algorithm outperformed the active contour random walker algorithm.…”

“…The mental condition of women suffering from ovarian cancer was analysed in Kodipalli & Devi, 2023 and Hegde and Kodipalli (2022)) and it was found that women with ovarian cancer have more mental problems compared to women without ovarian cancer. The novel variant of the CNN model to classify ovarian tumours as benign or malignant was proposed in (Kodipalli, Guha, et al, 2022) and the proposed model outperformed the state‐of‐the‐art architectures of 2014 winning architectures of ILSVRC. The novel single pipeline architecture for segmenting and classifying the tumours was proposed in (Kodipalli, Devi, et al, 2022) which not only outperformed in terms of time but also accuracy.…”